Contrast-Enhanced Image Analysis for MRI Based Multiple Sclerosis Lesion Segmentation

One of the most primary concern in Medical Image analyses is the detection of infected tumor in order to execute accurate treatment plan. In this paper, to segment lesions in Multiple Sclerosis (MS) pathology, we have investigated two preprocessing steps based on skull stripping (SS) and contrast enhancement (CE) which are two important steps for improving the quality rate of the MS lesion segmentation. After preprocessing step, a segmentation approach based on Expectation Maximization (EM) method have been applied to extract MS lesions. Qualitative and quantitative results of proposed method based on Dice score and Peak Signal to Noise Ratio was considered and tested on T2-F1air brain MR images.